The Physics of Chocolate

By Julianne Dalcanton | April 30, 2008 2:21 am

wedding_cake.jpg While deeply held feelings about string theory (“Genius!” “Total Bunk!”) may sometimes drive us apart, all of us can certainly get behind the theory that chocolate is a net good. However, in spite of its appeal as a tasty eatable (with or without bacon), it’s actually a bit of a pain to work with. If you’ve ever tried to use chocolate in its melted form, you’ve probably discovered that chocolate has a number of peculiarities that frequently thwart your best culinary efforts. For example, if your melted chocolate becomes contaminated with an errant drop of water, the chocolate siezes up. If you try to reharden chocolate that’s been melted (say, in making chocolate covered strawberries), you’re frequently left with a matte finish and crumbly texture that in no way resembles the dark glossy chocolate you began with.

The reasons for this should be familiar to any solid state physicist (or at least, they were to the one who made my wedding cake and first clued me in). Cocoa butter, one of the dominant ingredients in chocolate, contains several triglycerides that lock into a crystal form when cooled. However, there is not just one form that the triglycerides can lock into, but six of them (β(I) through β(VI)). Each successive form is more stable and has a higher melting point. Almost all commercial chocolate is in the β(V) form — from what I can tell, you only get to sample β(VI) in the afterlife, if you’ve been very, very good. When chocolate goes all wrong, it is usually a failure of the melted and cooled chocolate to recrystallize into the β(V) state. Similar problems can affect commercial chocolate suppliers as well, leading to chocolate that develops that unsightly chalky film we associate with old chocolate. Even previously stable β(V) chocolate can wind up with the same unsightly film after temperature fluctuations break down the crystal structure, or melt and reharden a thin layer on the surface. Given the commercial implications, there’s been some solid technical work on the structure of the magical β(V) form, which has been studied with x-ray diffraction using synchrotron radiation (more technical data here).

Given the above, when cooking with chocolate, one’s goal is to coax the cooled chocolate back into the β(V) form if one wants the end product to look glossy, be solid at room temperature, and have a nice crisp snap when bitten. The traditional mechanism for this is known as tempering (video here). Traditional tempering involves carefully controlling the temperature of the chocolate as it cools, so that the chocolate favors the preferred crystalline state. However, there is a vastly simpler mechanism, namely, seeding the crystal. If you take a lump of unmelted commercial chocolate, toss it into your bowl of melted chocolate, and stir for a bit, you’ll melt the new lump while cooling the melted chocolate. The cooling chocolate will then prefer the same crystal structure as the melting lump, such that when it hardens completely, you’ll find it in the luscious β(V) state.

PS. I can verify that the above works exactly as advertised. Last weekend I made the wedding cake above for the same solid state physicist who made mine a decade ago. (The cake was alternately described as looking like the Heatmiser‘s hair, Mordor, and Garrett Lisi’s E8 symmetry group, so you can imagine it was a pretty techie crowd). Making the thin chocolate sheets from which I cut the decorations, I got huge swaths of perfectly glossy chocolate. Occasionally, though, there’d be a small section with a matte surface, that was clearly a different crystalline form. Science. It works, bitches.

CATEGORIZED UNDER: Food and Drink, Science
  • Prof Smith

    Only string theory proponents actually appreciate chocolate. The opponents are, on the contrary, sourballs. The latter are sour, not sweet. See my disertation at Caltech, Like Chocolate for String theory,

    Sorry if it is too technical.

  • Pieter Kok

    Great post, and great cake! A trick I learned when water ruins your melted chocolate is to add a drop of vegetable or sunflower oil. I have no idea how it works on the level of the crystal structure, but it restores the creamy texture.

  • John R Ramsden

    Dunno if it helps with rehardening, but surprisingly a lot of chocolate recipes, both liquid and solid, also involve hot chili peppers.

    For example, there’s a recipe for Aztec Chili Hot Chocolate at:

  • j

    Pimping: It works bitches.
    Science: It works, bitches.

    The comma: it’s at least occasionally semantically important, bitches.

  • Ethan

    Sounds like ice-9 more than anything! Nice to know that Vonnegut did his homework before writing that one…

  • Jim Miles

    What a superb post. I hope we see more like it, in Sean’s absence. Great stuff!

  • Albatross

    For our wedding (nineteen years ago) we deliberately ordered a chocolate wedding cake, with chocolate frosting. We did this for three reasons: 1) Chocolate is good; 2) The chocolate wedding cake would so offend the eyes of both of our highly conservative families that they would be distracted from noticing any other flaws or mistakes at the wedding reception; 3) Chocolate is good.

    I am happy to say that the plan worked phenomenally, with our scandalized families nevertheless consuming chocolate-brown wedding cake, with many a critical whisper, and no attention paid whatsoever to any other problems at the reception.

    Also, we froze the leftover cake in triple plastic bags and tinfoil, and a year later when we consumed it on our first anniversary it was as delicious as on our wedding day. As a matter of fact, I believe I still have that cake, right about… here, on my hips.

  • Carl Brannen

    This was very useful and practical information. I had always imagined that the different crystalline states were due to the sugar either crystallizing or forming a glass.

    By the way, when one manufactures biodiesel from vegetable or animal fats, the triglycerides are broken into three oil molecules and a glycerol molecule. This is called “esterification” and is done with an alcohol. It takes three moles of alcohol to convert one mole of vegetable fat to 3 moles of biodiesel and a mole of glycerol. By weight, this works out to be about 10% ethanol or methanol to vegetable oil. So if they can get the algae process to make 100,000 gallons of oil per acre per year, and we replace all US gasoline consumption with biodiesel, there will still be a need for a huge ethanol industry.

  • Pingback: The Physics of Chocolate « Twisted One 151’s Weblog()

  • skeptic23

    Anyone who’s been on a college campus recently can testify that the ethanol industry in this country is thriving even without biodiesel.

  • Pingback: Salt in Water » mmmmmm, chocolate()

  • Ellipsis

    What a great post! I want to know more about the $beta(VI)$ state — is it not stable at STP? Or is there another reason why no one has figured out how to make and market it consistently?

  • Julianne

    Wikipedia claims that the β(VI) form takes weeks to form. Morover, as each crystal state is harder than the previous one, β(VI) might be too hard to actually eat in comfort.

  • Kennric

    Interestingly, when making chocolate ganache (that soft, creamy chocolaty filling in truffles), chocolate is melted and liquid deliberately added to it. It is a fascinating thing to do by hand, the first drops of liquid cause the chocolate to seize, and for a few minutes you’ll be sure you have just created a bowl of half-cured concrete. Then a magical phase transition occurs, and suddenly you are stirring perfectly smooth, shiny, delicious chocolate goo.

    The effect has to do with the liquid. Adding enough liquid eventually coats the microscopic bits of cocoa solid, allowing them to slide around one another instead of clinging fiercely. (A gross simplification, I am sure there is much interplay with the crystallizing triglycerides as well).

    I like to make mine with black tea and habenero peppers, for use in tiny hot truffles, but the traditional favorite is cream – though that is typically made by pouring hot cream over chipped chocolate, melting everything together at once and skipping the magical phase transition experience. Coffee is also excellent.

    Ganache is great for icing cakes when you want soft creamy frosting instead of hard chocolate shell, and you can fill strawberries with it to good effect.

  • Garrett

    That cake looks more like E7, but I approve (especially because of the strawberries).

  • Julianne

    The cake actually has a chocolate sour cream ganache as the filling. But, that’s just adding more fat (yay! more fat!) so it doesn’t go through the craziness that liquid-based ganaches do.

    And you’re completely right, Garrett. It’s exactly E7. I’ll try to dig up an overhead photo to confirm.

  • Lab Lemming

    I reckon that one reason that physics has a higher profile than chemistry today is that whenever someone demonstrates cool chemistry, they call it physics instead.

    That being said, it is an awesome cake.

  • John

    Beta-6 crystals are available commercially so you can make your own beta-6 chocolate without waiting for the afterlife.

  • Julianne

    Aha! It appears that β(VI) might be what you get in couverture chocolate, which is a very crisp chocolate that is frequently used for dipping.

  • Pingback: delgrosso dot com » Physics of chocolate()

  • minstrel hussain boy

    i’m a chocolatier. i have a couple of tempering machines which i use with great success. the “seed” method will work fairly well. but, if the chocolate you are using has broken temper you will need to bring the whole shebang up to 118 (f), hold that while working the chocolate for fifteen minutes. cool, while still working, down to 92, heat back up to 99 (all this while working the chocolate which the temperers do by having the bowl spin past a dasher) and then, cool slowly back down to between 90 and 92 while adding the “seed” to guide the formation of desirable crystals.

    the easiest way is to never break the temper. the best way to do this is with a heating pad, like for a bad back (which most chocolatiers have anyway) and an instant read (preferably laser so you don’t have to touch the chocolate) thermometer. while melting, and stirring, never, ever, at all, allow the chocolate to get above 95.

    in a micro wave you can melt small chunks of tempered chocolate by nuking on high for 30 seconds. stir thoroughly, repeat for 30 seconds. stir. then reduce the time to 15 seconds, twice. then in 10 second increments, stirring, between each set. this should melt chocolate well enough to do dipping or painting without breaking the temper.

    anyone who is needing advice or instruction may feel absolutely free to click through to my homepage. drop a comment or send me an email and i will be happy to try and help you understand the beauty and voodoo of this beautiful and psychoactive treat.

  • Julianne

    Thanks for details from a true expert. The reason I may have had success with the seed method is that I have a really low power microwave. When I melt chocolate, I run it on defrost (which turns the microwave on and off intermittently, and stops a number of times and beeps at you until you restart it). Combined with the low wattage, I usually never actually bring the chocolate to full melt, and just stir by hand to get the last chunks to melt. I might therefore never be breaking temper.

  • Chris W.

    Perhaps you’ll soon get more comments from the chocolatiers at Lake Champlain Chocolates; see the comment on this post on the company blog.

  • Pingback: Schokolade im Klimawandel und was Wissenschaftsjournalisten mit abgetrennten Fingern machen « Begrenzte Wissenschaft()

  • Kishan Yerubandi

    Unfortunately, not all of us can get behind the claim that chocolate is a net good. In fact, most of the world’s cocoa is picked by the hands of slaves in western Affrica. Yes, documented slavery exists in the year 2008.

    To catch a glimpse of children who are abducted and enslaved in cocoa farms in the Ivory Coast, I recommend this streaming documentary:

    I personally have boycotted chocolate.

  • The Almighty Bob

    What about FairTrade chocolate, Kishan?

  • Alex Cull

    The point is, though, as chocolate is composed of hydrocarbons, can we use it as a fuel source? Would a car running on dark chocolate, for instance, be more economical than one running on milk or white chocolate? Personally, I think this would be a bad idea, on balance; if the cocoa crop was turned into biofuel, we’d see a worldwide shortage of chocolate, and then there’d be trouble. :-)

  • Claire C Smith


    I was thinking about Truffles.

    Is it like, you get a solid state piece of chocolate, dunk it in liquid chocolate, like from B(I) through to B(VI) state, then as it where, you then dunk that when it’s still wet into powdered chocolate. This type of Truffle would then resemble a AGB star and a RGB star, but both are big, chocolatey, and about to die (be eaten), but the AGB Truffles are bigger chocs, and thus more luminous (hundreds and thousands?), and prone to screwing around with the amount of taste you get?

    Chocolate science: It works, bitches.


  • Kevin Schnitzius

    Glossy finish or not, it still gets eaten and ends up you-know-where…

  • Pingback: Creative Science Camp for Kids - LEGO Camp » Life Goals Articles from SuperViva()

  • Chocolate

    Chocolate is my passion, but this sounds a bit too complicated to me. :-)

  • Victor

    Albatross; check the ingredients, other than cocoa, in the chocolate you are buying. If the chocolate has 20% cocoa that leaves a lot of room for junk such as sugar, emulsifiers, etc. and theres the danger. If the chocolate is 85% cocoa there is far less room for additives. It then has a bitter but delicious addictive flavour.

  • Pingback: iChoc: New Chocolate Factory Operated Entirely by iPhone | Discoblog | Discover Magazine()

  • Pingback: Women in Space — We …um… salute you | Cosmic Variance | Discover Magazine()

  • Pingback: White lovers and eggy milkshakes « Lagom()


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Cosmic Variance

Random samplings from a universe of ideas.

See More

Collapse bottom bar